[libc][obvious] Correctly hoist mask out of the loop

[jhuber6]

Summary

Summry:
This was accidentally dropped from a previous patch following a rebase.
Fix it to where it's consistent.

libc/src/__support/RPC/rpc.h

num_sends = is_process_gpu() ? lane_value(size, id)
                                  : max(lane_value(size, id), num_sends);
     uint64_t len =
         lane_value(size, id) > sizeof(Buffer::data) - sizeof(uint64_t)
             ? sizeof(Buffer::data) - sizeof(uint64_t)
             : lane_value(size, id);
     inline_memcpy(&buffer->data[1], lane_value(src, id), len);
   });
   uint64_t idx = sizeof(Buffer::data) - sizeof(uint64_t);
-  while (gpu::ballot(process.get_packet(index).header.mask, idx < num_sends)) {
+  uint64_t mask = process.get_packet(index).header.mask;
+  while (gpu::ballot(mask, idx < num_sends)) {
     send([=](Buffer *buffer, uint32_t id) {
       uint64_t len = lane_value(size, id) - idx > sizeof(Buffer::data)
                          ? sizeof(Buffer::data)
                          : lane_value(size, id) - idx;
       if (idx < lane_value(size, id))
         inline_memcpy(buffer->data, advance(lane_value(src, id), idx), len);
     });
     idx += sizeof(Buffer::data);